The increasing availability of large healthcare databases is fueling an intense debate on whether real-world data should play a role in the assessment of the benefit-risk of medical treatments. In ...many observational studies, for example, statin users were found to have a substantially lower risk of cancer than in meta-analyses of randomized trials. Although such discrepancies are often attributed to a lack of randomization in the observational studies, they might be explained by flaws that can be avoided by explicitly emulating a target trial (the randomized trial that would answer the question of interest). Using the electronic health records of 733,804 UK adults, we emulated a target trial of statins and cancer and compared our estimates with those obtained using previously applied analytic approaches. Over the 10-yr follow-up, 28,408 individuals developed cancer. Under the target trial approach, estimated observational analogs of intention-to-treat and per-protocol 10-yr cancer-free survival differences were -0.5% (95% confidence interval (CI) -1.0%, 0.0%) and -0.3% (95% CI -1.5%, 0.5%), respectively. By contrast, previous analytic approaches yielded estimates that appeared to be strongly protective. Our findings highlight the importance of explicitly emulating a target trial to reduce bias in the effect estimates derived from observational analyses.
To emulate three target trials: single treatment vs. no treatment, joint treatment vs. no treatment, and head-to-head comparison of two treatments, we explain how to estimate the observational ...analogs of intention-to-treat and per-protocol effects, using hazard ratios and survival curves. For per-protocol effects, we describe two methods for adherence adjustment via inverse-probability weighting.
Prospective observational study using electronic medical records of individuals aged 55–84 with coronary heart disease from >500 practices in the United Kingdom between 2000 and 2010.
The intention-to-treat mortality hazard ratio (95% confidence interval) was 0.90 (0.84, 0.97) for statins vs. no treatment, 0.88 (0.73, 1.06) for statins plus antihypertensives vs. no treatment, and 0.91 (0.77, 1.06) for atorvastatin vs. simvastatin. When censoring nonadherent person-times, the per-protocol mortality hazard ratio was 0.74 (0.64, 0.85) for statins vs. no treatment, 0.55 (0.35, 0.87) for statins plus antihypertensives vs. no treatment, and 1.13 (0.88, 1.45) for atorvastatin vs. simvastatin. We estimated per-protocol hazard ratios for a 5-year treatment using different dose-response marginal structural models and standardized survival curves for each target trial using intention-to-treat and per-protocol analyses.
When randomized trials are not available or feasible, observational analyses can emulate a variety of target trials.
Randomized trials have shown that initiating breast cancer screening between ages 50 and 69 years and continuing it for 10 years decreases breast cancer mortality. However, no trials have studied ...whether or when women can safely stop screening mammography. An estimated 52% of women aged 75 years or older undergo screening mammography in the United States.
To estimate the effect of breast cancer screening on breast cancer mortality in Medicare beneficiaries aged 70 to 84 years.
Large-scale, population-based, observational study of 2 screening strategies: continuing annual mammography, and stopping screening.
U.S. Medicare program, 2000 to 2008.
1 058 013 beneficiaries aged 70 to 84 years who had a life expectancy of at least 10 years, had no previous breast cancer diagnosis, and underwent screening mammography.
Eight-year breast cancer mortality, incidence, and treatments, plus the positive predictive value of screening mammography by age group.
In women aged 70 to 74 years, the estimated difference in 8-year risk for breast cancer death between continuing and stopping screening was -1.0 (95% CI, -2.3 to 0.1) death per 1000 women (hazard ratio, 0.78 CI, 0.63 to 0.95) (a negative risk difference favors continuing). In those aged 75 to 84 years, the corresponding risk difference was 0.07 (CI, -0.93 to 1.3) death per 1000 women (hazard ratio, 1.00 CI, 0.83 to 1.19).
The available Medicare data permit only 8 years of follow-up after screening. As with any study using observational data, the estimates could be affected by residual confounding.
Continuing annual breast cancer screening past age 75 years did not result in substantial reductions in 8-year breast cancer mortality compared with stopping screening.
National Institutes of Health.
Abstract
Background
Previous case-control studies have reported a strong association between statin use and lower cancer risk. It is unclear whether this association reflects a benefit of statins or ...is the result of design decisions that cannot be mapped to a (hypothetical) target trial (that would answer the question of interest).
Methods
We outlined the protocol of a target trial to estimate the effect of statins on colorectal cancer incidence among adults with low-density lipoprotein (LDL) cholesterol below 5 mmol/L. We then emulated the target trial using linked electronic health records of 752 469 eligible UK adults (CALIBER 1999–2016) under both a cohort design and a case-control sampling of the cohort. We used pooled logistic regression to estimate intention-to-treat and per-protocol effects of statins on colorectal cancer, with adjustment for baseline and time-varying risk factors via inverse-probability weighting. Finally, we compared our case-control effect estimates with those obtained using previous case-control procedures.
Results
Over the 6-year follow-up, 3596 individuals developed colorectal cancer. Estimated intention-to-treat and per-protocol hazard ratios were 1.00 (95% confidence interval CI: 0.87, 1.16) and 0.90 (95% CI: 0.71, 1.12), respectively. As expected, adequate case-control sampling yielded the same estimates. By contrast, previous case-control analytical approaches yielded estimates that appeared strongly protective (odds ratio 0.57, 95% CI: 0.36, 0.91, for ≥5 vs. <5 years of statin use).
Conclusions
Our study demonstrates how to explicitly emulate a target trial using case-control data to reduce discrepancies between observational and randomized trial evidence. This approach may inform future case-control analyses for comparative effectiveness research.
Patients with cancer who use statins appear to have a substantially better survival than nonusers in observational studies. However, this inverse association between statin use and mortality may be ...due to selection bias and immortal-time bias.
To emulate a randomized trial of statin therapy initiation that is free of selection bias and immortal-time bias.
We used observational data on 17 372 patients with cancer from the Surveillance, Epidemiology, and End Results (SEER)-Medicare database (2007-2009) with complete follow-up until 2011. The SEER-Medicare database links 17 US cancer registries and claims files from Medicare and Medicaid in 12 US states. We included individuals with a new diagnosis of colorectal, breast, prostate, or bladder cancer who had not been prescribed statins for at least 6 months before the cancer diagnosis. Individuals were duplicated, and each replicate was assigned to either the strategy "statin therapy initiation within 6 months after diagnosis" or "no statin therapy initiation." Replicates were censored when they stopped following their assigned strategy, and the potential selection bias was adjusted for via inverse-probability weighting. Hazard ratios (HRs), cumulative incidences, and risk differences were calculated for all-cause mortality and cancer-specific mortality. We then compared our estimates with those obtained using the same analytic approaches used in previous observational studies.
Statin therapy initiation within 6 months after cancer diagnosis.
Cancer-specific and all-cause mortality using SEER-Medicare data and data from previous studies.
Of the 17 372 patients whose data were analyzed, 8440 (49%) were men, and 8932 (51%) were women (mean SD age, 76.4 7.4 years; range, 66-115 years). The adjusted HR (95% CI) comparing statin therapy initiation vs no initiation was 1.00 (0.88-1.15) for cancer-specific mortality and 1.07 (0.93-1.21) for overall mortality. Cumulative incidence curves for both groups were almost overlapping (the risk difference never exceeded 0.8%). In contrast, the methods used by prior studies resulted in an inverse association between statin use and mortality (pooled hazard ratio 0.69).
After using methods that are not susceptible to selection bias from prevalent users and to immortal time bias, we found that initiation of therapy with statins within 6 months after cancer diagnosis did not appear to improve 3-year cancer-specific or overall survival.
Metformin users appear to have a substantially lower risk of cancer than nonusers in many observational studies. These inverse associations may be explained by common flaws in observational analyses ...that can be avoided by explicitly emulating a target trial.
We emulated target trials of metformin therapy and cancer risk using population-based linked electronic health records from the UK (2009-2016). We included individuals with diabetes, no history of cancer, no recent prescription for metformin or other glucose-lowering medication, and hemoglobin A1c (HbA1c) <64 mmol/mol (<8.0%). Outcomes included total cancer and 4 site-specific cancers (breast, colorectal, lung, and prostate). We estimated risks using pooled logistic regression with adjustment for risk factors via inverse-probability weighting. We emulated a second target trial among individuals regardless of diabetes status. We compared our estimates with those obtained using previously applied analytic approaches.
Among individuals with diabetes, the estimated 6-year risk differences (metformin - no metformin) were -0.2% (95% CI = -1.6%, 1.3%) in the intention-to-treat analysis and 0.0% (95% CI = -2.1%, 2.3%) in the per-protocol analysis. The corresponding estimates for all site-specific cancers were close to zero. Among individuals regardless of diabetes status, these estimates were also close to zero and more precise. By contrast, previous analytic approaches yielded estimates that appeared strongly protective.
Our findings are consistent with the hypothesis that metformin therapy does not meaningfully influence cancer incidence. The findings highlight the importance of explicitly emulating a target trial to reduce bias in the effect estimates derived from observational analyses.
To compare the risk of coronavirus disease 2019 (COVID-19) outcomes by antiretroviral therapy (ART) regimens among men with HIV.
We included men with HIV on ART in the Veterans Aging Cohort Study ...who, between February 2020 and October 2021, were 18 years or older and had adequate virological control, CD4 + cell count, and HIV viral load measured in the previous 12 months, and no previous COVID-19 diagnosis or vaccination.
We compared the adjusted risks of documented severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, COVID-19-related hospitalization, and intensive care unit (ICU) admission by baseline ART regimen: tenofovir alafenamide (TAF)/emtricitabine (FTC), tenofovir disoproxil fumarate (TDF)/FTC, abacavir (ABC)/lamivudine (3TC), and other. We fit pooled logistic regressions to estimate the 18-month risks standardized by demographic and clinical factors.
Among 20 494 eligible individuals, the baseline characteristics were similar across regimens, except that TDF/FTC and TAF/FTC had lower prevalences of chronic kidney disease and estimated glomerular filtration rate <60 ml/min. Compared with TAF/FTC, the estimated 18-month risk ratio (95% confidence interval) of documented SARS-CoV-2 infection was 0.65 (0.43, 0.89) for TDF/FTC, 1.00 (0.85, 1.18) for ABC/3TC, and 0.87 (0.70, 1.04) for others. The corresponding risk ratios for COVID-19 hospitalization were 0.43 (0.07, 0.87), 1.09 (0.79, 1.48), and 1.21 (0.88, 1.62). The risk of COVID-19 ICU admission was lowest for TDF/FTC, but the estimates were imprecise.
Our study suggests that, in men living with HIV, TDF/FTC may protect against COVID-19-related events. Randomized trials are needed to investigate the effectiveness of TDF as prophylaxis for, and early treatment of, COVID-19 in the general population.
Researchers are often interested in using observational data to estimate the effect on a health outcome of maintaining a continuous treatment within a prespecified range over time, for example, ..."always exercise at least 30 minutes per day." There may be many precise interventions that could achieve this range. In this article, we consider representative interventions. These are special cases of random dynamic interventions: interventions under which treatment at each time is assigned according to a random draw from a distribution that may depend on a subject's measured past. Estimators of risk under representative interventions on a time-varying treatment have previously been described based on g-estimation of structural nested cumulative failure time models. In this article, we consider an alternative approach based on inverse probability weighting (IPW) of marginal structural models. In particular, we show that the risk under a representative intervention on a time-varying continuous treatment can be consistently estimated via computationally simple IPW methods traditionally used for deterministic static (i.e., "nonrandom" and "nondynamic") interventions for binary treatments. We present an application of IPW in this setting to estimate the 28-year risk of coronary heart disease under various representative interventions on lifestyle behaviors in the Nurses' Health Study. Supplementary materials for this article are available online.